The Landscape of Maritime Piracy and the Limits of Statistical Prediction

Abstract

Maritime piracy has been on the rise for many years. Although public data on piracy attacks and piracy networks is limited, we have attempted to apply some new methodological tools to determine as far as possible the properties of these complex systems. In particular, by employing extremal statistics and following the work of Clauset and Gleditsch on land-based terrorist organizations, we attempt to characterize the relationship between the severity and frequency of maritime piracy attacks. In addition, utilizing social network analysis, we have aimed at accurately visualizing and representing key characteristics of maritime piracy as well as discerning non-obvious or emergent patterns of behavior and organizational structure specific to piracy networks as defined by their geographic location in Appendix I (i.e., West Africa, East Africa, and The Straits of Malacca and Lombok). In Appendix II, we plot the distribution of attacks by region, showing how counter-intuitive the patterns are and how profoundly they differ from the land-based patterns uncovered by Clauset et al. Finally, in Appendix III, Britten Schear reviews the unique historical factors which characterize Somali piracy.